Home
Rational® Functional Tester
Test Author Guide
This guide describes how to create test scripts and enhances them by applying different test elements such as dataset, variables, and verification points.
Testing in the UI Test perspective
When you develop web applications, Windows-based applications, or mobile applications, you can use IBM® Rational® Functional Tester to create functional tests for these applications. You must first record the tests and then use Rational Functional Tester to run the tests before you can view the test results.
Testing web applications
You can test web applications in Rational® Functional Tester by using the industry-standard browsers such as Google Chrome, Apple Safari, Mozilla Firefox, and Microsoft Edge. You can record Web UI tests and then play back the tests to evaluate the test results. You can also test the web applications on dual monitors when you extend the display of your computer to a secondary monitor.
Editing Web UI tests
With the test editor, you can view and customize a Web UI test.
Optimizing Web UI load testing for scalability
Scalability denotes the number of virtual users that can be emulated on an agent machine to generate load. In a Web UI load test, each virtual user requires one instance of a web browser to be run. Because each instance of a browser consumes key computer resources such as memory, CPU and Network data, you must tune a few of the parameters to comfortably run appropriate number of browser (virtual users) on an agent machine.

Rational® Functional Tester
- Release Notes
  This document contains information about what's new, installation instructions, known problems in and contact information of IBM Customer Support.
- System Requirements
  This document includes information about hardware and software requirements for IBM® Rational® Functional Tester.
- Getting Started Guide
  This guide provides an overview of Rational® Functional Tester. You can find the task flows to get you started with Rational Functional Tester. This guide is intended for new users.
- Tutorials
  This section contains the tutorials which explains the main features of Rational® Functional Tester.
- Samples
  This section describes about the sample project which can be used with Rational® Functional Tester to test the functionality of an application.
- Administrator Guide
  This guide describes how to install IBM® Rational® Functional Tester software. After you install the software, you can perform administration tasks such as license configuration and integration with other products. This guide is intended for administrators.
- Test Author Guide
  This guide describes how to create test scripts and enhances them by applying different test elements such as dataset, variables, and verification points.
  - Testing in the UI Test perspective
    When you develop web applications, Windows-based applications, or mobile applications, you can use IBM® Rational® Functional Tester to create functional tests for these applications. You must first record the tests and then use Rational Functional Tester to run the tests before you can view the test results.
    - Testing web applications
      You can test web applications in Rational® Functional Tester by using the industry-standard browsers such as Google Chrome, Apple Safari, Mozilla Firefox, and Microsoft Edge. You can record Web UI tests and then play back the tests to evaluate the test results. You can also test the web applications on dual monitors when you extend the display of your computer to a secondary monitor.
      - Creating Web UI tests
        See the different ways to record Web UI tests. After the recording stops, the tests are generated automatically.
      - Editing Web UI tests
        With the test editor, you can view and customize a Web UI test.
        Inserting browser navigation actions in a Web UI test
        While editing a Web UI test, you can add or insert browser navigation actions, such as Back, Forward, Close, Maximize, Refresh, Restore, GoToUrl, and several others.
        Defining a variable to run a test with a selected browser
        If you run a test from the command line, as part of a schedule from IBM® Rational® Performance Tester, or from IBM Rational Quality Manager, you can define a variable in the test to specify the web browser to use for running the test. The reserved name for the variable is RTW_WebUI_Browser_Selection. After defining the variable, if you run the test from test workbench, the browser selected in the Run configuration dialog box takes precedence over the browser specified in the variable.
        Creating verification points in a test
        You can create verification points for any object properties, such as label, color, and count, and you can verify that an object property is enabled, that it has focus, whether it is clickable, and so on.
        Creating verification points for alert, confirm, or prompt dialog box
        You can create verification points in alert, confirm, or prompt dialog box of a recorded Web UI test script to validate text messages during playback.
        Assigning a test variable to an objects property
        You can assign a new value to a test variable and set it to a Mobile objects property.
        Creating verification points from the SmartShot View
        You can create some verification points in a test with simplified scripts by using the SmartShot View. A verification point can be added for an object or created for the properties of the object.
        Adding a loop
        To run a test or part of a test repeatedly for a specified number of times or duration, add a loop to the test. You can add a loop to the Launch Application node or In Application node.
        Adding conditional logic to tests
        You can insert IF-ELSE logic around portions of a test to make those portions run if a specific condition is met.
        Modularizing test script
        After you record a test, you can modularize or split the test into multiple test segments with different nodes. With the test-splitting capability, you can record a relatively long scenario with many functional steps against an application and then, in the editor, modify the target apps. Then you can generate multiple tests from a single recording that you can replay in a different order with a schedule. You can also insert in a test such elements as transactions, loops, IF-THEN conditions or random selectors.
        Optimizing the playback time of a test
        When you play back a Web UI test, by default, the test collects a lot of data. This data includes screen shots, response time breakdown data, and highlighted UI controls. The size of a test depends upon the data it contains; playback time increases as the size of a test increases. To optimize the playback time, you can choose to collect only the data which you need.
        Optimizing Web UI load testing for scalability
        Scalability denotes the number of virtual users that can be emulated on an agent machine to generate load. In a Web UI load test, each virtual user requires one instance of a web browser to be run. Because each instance of a browser consumes key computer resources such as memory, CPU and Network data, you must tune a few of the parameters to comfortably run appropriate number of browser (virtual users) on an agent machine.
        Actions from the SmartShot View
        The SmartShot View displays the screen captures that were captured during the recording of the application under test. Use this view to display and select user interface (UI) elements and optionally add verification points to the test script.
        Improving test script robustness
        Sometimes the recorded steps in a test cannot be recognized when the test is played back, thus leading to test failures or errors. To avoid these object recognition issues, you can change the manner in which objects are identified, you can use object locator conditions, or apply responsive design conditions, or add a list of preferred properties to be used during the test run. You can also perform asynchronous user actions for the steps playback to improve the automated test reliability. That way, you improve test robustness and improve the chances that the test can be included in an automated testing process.
        Putting test assets under source control
        Use version-control software, such as Rational® ClearCase® or Rational Team Concert™, to put the test assets under source control.
        Using guided healing and self-healing features to update test scripts
        As applications undergo changes during regression testing, application GUI is frequently modified and the regression tests must be updated to identify possible issues, and to ensure that the new code has not affected other parts of the software. Some changes can be easily detected such as edit box labels, but others can be more difficult to identify, for example, the UI control properties. To help testers to identify those changes in the application and to fix the Web UI tests accordingly, more intelligence has been added to test playback with the guided healing and self-healing features.
        Substituting the URL of an application by using Datasets
        If you have the same Web UI application deployed on different servers corresponding to different phases of a project (development, production...), or different versions of the application for example, you might want to execute the same Web UI test on all these servers at once. Initially, one way to do so was to create web applications for all these different servers, then create a copy of the same test and add it to a compound test that you would run. From version 9.2.1 of the product, you can use a more straightforward procedures. To switch from one server to another, you can just edit the URL of the application and run the test on a different server. To execute tests on different environments at once, you can substitute the URL from a dataset so that you can run the same test in a loop on different servers.
        Extension of application URL in Web UI tests
        Validating images and user interface elements by using the image property
        When you play back the Web UI test, you can validate the images and user interface elements by the using the image property. After you define a test step to use image property, during playback, Rational® Functional Tester locates and compares the selected image with the target objects in the web pages and finds the best matching image.
        Defining the Image property as object identifier for Web UI tests
        Before you start recording Web UI tests, you can define the image property as the object identifier. When image property is set as the object identifier, then the objects in the web application are recognized and captured as images while recording a Web UI test.
        Applying guided healing feature for tests identified by the image property
        When you play back any Web UI tests that use the image property to match target with actual images, the tests fail if there is a mismatch. You can use the guided healing feature that provides the capability of finding the next best matching image so that the tests can pass.
        Creating a custom JavaScript code in a Web UI test
        You can use the JavaScript Custom Code option to create your own code and add that as a step in the Web UI test.
        Extending Web UI tests
        You can perform certain tasks to extend the Web UI tests that include advanced capabilities such as adding JavaScript code to tests, replacing a JavaScript in a test script, and creating a custom Java code.
        Providing tests with variable data (datasets)
        You can produce more realistic tests by changing them to use datasets. During execution, a test that uses a dataset replaces a value in the recorded test with variable test data that is stored in the dataset. This substitution allows each virtual user to generate a different request to the server.
      - Frequently asked questions
        This section provides answers to frequently asked questions about the IBM® Rational® Functional TesterWeb UI extension.
      - Troubleshooting Web UI tests
        While working with Web UI tests, you might encounter problems that you can easily troubleshoot. The topics in this section address some of these problems.
    - Testing mobile applications
      You can test the Android or iOS mobile applications by recording and playing back mobile tests on Android devices or emulators, iOS mobile devices or simulators by using Rational® Functional Tester.
    - Testing Windows desktop applications
      You can configure the Windows applications to record and play back tests on the Windows desktop applications. You can also test the Windows applications on dual monitors when you extend the display of your computer to a secondary monitor.
    - Recording SAP tests
      When you record a test, the test creation wizard records your interactions with the SAP server, generates a test from the recording, and opens the test for editing. You can record tests from the SAP GUI. You can also record SAP batch input tests that can be used to produce a heavy load on the server while minimizing the processing requirements for virtual testers.
    - Working with Selenium or Appium tests
      You can manage your Selenium or Appium Java™ tests from the UI Test perspective, where you can import and modify the tests, add them to compound tests, and run them. You can also create new Appium Java tests from the Perfecto perspective on Windows™ and Mac OS, add the tests to compound tests, and run them. Perfecto is not currently supported on Linux®.
    - Compound tests
      You can create compound tests to help you organize smaller tests into scenarios that can then be run end-to-end. You can combine tests from different extensions to achieve end-to-end flow.
    - Accelerated Functional Tests
      When you create Web UI, mobile, or Windows tests, you can use the Accelerated Functional Test (AFT) feature in Rational® Functional Tester to distribute test effort during the play back of the recorded tests.
    - Working with keywords
      In Rational® Functional Tester, you can record and then associate Test Workbench tests with the keywords that are created in IBM® Rational Quality Manager. The keywords are used in manual tests in Rational Quality Manager. When you run a manual test that contains a keyword associated with a Rational Test Workbench test, the Test Workbench test is run on the workstation where the adapter has been started.
  - Testing in the Functional Test perspective
    When you develop Java, HTML, or Terminal-based applications, you can use IBM® Rational® Functional Tester to create traditional functional tests for these applications. You must first record the tests and then use Rational Functional Tester to run the tests before you can view the test results.
  - Extending the UI Test perspective with custom code
    You can use custom code to extend the default Web UI testing capabilities. You can write custom Java™ code and call the code from the test. You can also specify that results from the tests that are affected by your custom code to be included in reports.
  - Extending the Functional Test perspective
    You can use custom code to extend the default functional testing capabilities. You can write custom code and call the code from the test. You can also specify that results from the tests that are affected by your custom code to be included in reports.
  - Experimental Features
    Rational® Functional Tester contains a set of pre-releases that enable you to test various experimental features. These experimental features, while still in progress, are introduced early in the release to seek your feedback on its overall functionality and performance. You can play around with these capabilities before they are made available and supported as a general feature in an upcoming release.
- Test Execution Specialist Guide
  This guide describes how to automate the playback of the tests by using different methods such as Docker, IBM® Cloud Private, Jenkins and so on. The test results can be pushed to Rational® Test Control Panel.
- Test Manager Guide
  This guide describes how to keep track of the performance of the application by evaluating the test results. Following topics cover how you can work with the test results.
- Troubleshooting
  This guide describes how to analyze and resolve some of the common problems that you might encounter while you work with Rational® Functional Tester.
- Reference Guide
  This guide describes the additional topics to gain more knowledge about Rational® Functional Tester.
- All Guides (PDF)
  You can find the link to view or download the content of the release notes, system requirements, and persona guides of IBM® Rational® Functional Tester as a PDF file.
- Security Considerations
  You can take certain actions to ensure that your installation is secure, customize your security settings, and set up user access controls.
- Notices
  This document provides information about copyright, trademarks, terms and conditions for the product documentation.

Optimizing Web UI load testing for scalability

Scalability denotes the number of virtual users that can be emulated on an agent machine to generate load. In a Web UI load test, each virtual user requires one instance of a web browser to be run. Because each instance of a browser consumes key computer resources such as memory, CPU and Network data, you must tune a few of the parameters to comfortably run appropriate number of browser (virtual users) on an agent machine.

About this task

Here is the sample configuration of a computer that was used to determine the memory consumption:

Table 1. Sample configuration
Operating System	Red Hat Enterprise Linux 6.6
System main memory	4 GB
Free memory before test run	~3 GB
Firefox (empty) memory consumption	130 MB
Firefox (medium-size app) main memory	200 MB

With this configuration, about 15-20 (300 MB / 200 MB) browser instances or virtual users can be emulated on on agent machine. Note, the browser memory consumption may vary based on the dom size of the web applications.

Tuning parameters

By default, the playback process of the Web UI test creates a thread pool of size 10 threads. This behavior affects the execution time when more than 10 virtual users per agent machine are emulated. This limit can be improved by passing the =-DrptDynamicThreads parameter to the agent machines.

Also, the Web UI test playback is a java process with maximum memory limit set to 512MB through the –Xmx argument. This parameter might yield more memory for browser instantiations.

Procedure

To optimize Web UI load test:

In the Test Navigator view, double-click the location assets for optimization.
In the General Properties tab, double-click RPT_VMARGS row.
If RPT_VMARGS parameter is not available, click Add.
In Property Name, specify RPT_VMARGS and in Property Value specify -DrptDynamicThreads -Xmx512m.
Click OK and save the change.